Real-Time Music Visualization on the iPhone GPU

https://deezer.io/real-time-music-visualization-on-the-iphone-gpu-579d631272d3

by birslip • 9 years ago

94 15 9 years ago

mpalmer• 9 years ago

This looks interesting but it's pretty hard to judge the quality of real time audio visualization with no audio.

I wrote a very similar prototype a while ago. Here's a video - https://www.youtube.com/watch?v=oYZEDSzX7MU&feature=youtu.be

I didn't realise at the time, but the X axis is not in logarithmic scale, so most of the audible frequencies are bunched on the left hand side.

stiGGG• 9 years ago

Nice, do you share the code?

rlods• 9 years ago

you're right, thank you for the feedback, we'll try to add video samples with royalty free audio content.

taylodl• 9 years ago

I'd be interested in seeing the power load of these visualizations - are they a battery drainer?

birslip• 9 years ago

There is definitely an impact on the battery. The FFT part is well optimised (thanks to the Accelerate framework). What is costly is mainly the shader... and it really depends on the shader. A very simple shader is usually less than 5% of the GPU usage (according the Xcode Instruments, and on a recent iPhone). However if you do a lot in your shader it's going to be much more costly.

rocky1138• 9 years ago

At first I was thinking "this isn't a big deal, Atari Jaguar CD did this back in 1995 on way less hardware," but then I realised it was a hackathon project and now I'm really impressed at how far they got in that short frame of time!

pmlnr• 9 years ago

I miss Winamp effects as well.

parkersweb• 9 years ago

Yeah - we need a Geiss port :) http://www.geisswerks.com/geiss/

acoye2• 9 years ago

Is the code available somewhere? It doesn’t look like to be open-sourced.

thedjinn2• 9 years ago

Pretty boring if it's just FFT based. I'd love to see music visualizations make use of more sophisticated musical feature detection, like polyphonic note recognition or tempo/rhythm analysis.

acoye2• 9 years ago

Hey, I'm Adrian. I did the 'More colors' shader.

Given the nature of the project, a hackathon, sticking to the last computed FFTs and a notion of time was all we need to play with nice shader effects.

I'm currently playing with tensorflow on my spare time. I'd like to use a neural net to find interesting features of a track and inputting this to a shader. That's what I would do if I had another go at it.

stillkicking• 9 years ago

Cool work, but note that a 1024 point FFT isn't really sufficient to visualize music. At 44.1kHz, the lowest frequency you can resolve is 43Hz. Anything below that will be folded into your DC component. Your bass response is mostly going to be overall volume, rather than beat.

Real-Time Music Visualization on the iPhone GPU

Related Stories